I committed a fix for this issue in trunk.  java.util.InetAddress is
not referenced anywhere in the source now.  Could one of our GAE users
try out the latest build in a GAE deployment and let us know how it
goes?

Thanks,

Les

On Fri, Dec 18, 2009 at 11:52 AM, Les Hazlewood <[email protected]> wrote:
> After looking at the code, I think only a 'getHost' method should be
> used.  This maintains a 1:1 association with existing calls.
> Otherwise we'd have to duplicate those methods almost everywhere the
> InetAddress was accessed previously - a less than ideal refactoring
> exercise I think.
>
> If additional information is needed for more special use-cases, custom
> interfaces and/or token implementations can easily be used.
>
> - Les
>
> On Fri, Dec 18, 2009 at 11:09 AM, Les Hazlewood <[email protected]> wrote:
>> I was going to create a HostAuthenticationToken interface to replace
>> the InetAuthenticationToken with methods:
>>
>> getHostname() : String
>> getHostAddress() : String
>>
>> The implementations would mirror the ServletRequest's getRemoteHost()
>> and getRemoteAddr() methods respectively (but could be used outside of
>> a Servlet environment of course).
>>
>> For XMPP, if this doesn't make sense, you could always just create a
>> custom UsernamePasswordToken subclass that implements an
>> XmppAuthenticationToken interface that returns the JID or whatever
>> else you may need.
>>
>> Let me know if anyone has any objections or further ideas.
>>
>> Regards,
>>
>> Les
>>
>> On Fri, Dec 18, 2009 at 9:00 AM, Jason <[email protected]> wrote:
>>> cool - this will also mean I can use an XMPP jid instead of an inetAddress
>>> here which would be very handy.
>>>
>>> Les Hazlewood wrote:
>>>>
>>>> We didn't receive any objections to this change, so I've created an
>>>> issue to track it:
>>>>
>>>> https://issues.apache.org/jira/browse/SHIRO-121
>>>>
>>>> Regards,
>>>>
>>>> Les
>>>>
>>>> On Thu, Dec 17, 2009 at 3:28 PM, Tauren Mills <[email protected]> wrote:
>>>>>
>>>>> I would love to have GAE compatibility and won't cry over the loss, I say
>>>>> go
>>>>> for it!
>>>>>
>>>>> Tauren
>>>>>
>>>>>
>>>>> On Fri, Dec 11, 2009 at 10:03 AM, Jeremy Haile <[email protected]> wrote:
>>>>>
>>>>>> I love that idea!  Using InetAddress is mostly just an annoyance I
>>>>>> think!
>>>>>>
>>>>>>
>>>>>>
>>>>>> Les Hazlewood wrote:
>>>>>>
>>>>>>> I was thinking about Google App Engine and then I remembered this
>>>>>>> thread and wanted to bump it up for discussion.
>>>>>>>
>>>>>>> What does everyone think about changing all references from
>>>>>>> InetAddress to a String?  There are 181 places in code this is
>>>>>>> referenced, so its not a trivial change  (but not difficult), and
>>>>>>> would break some backwards-compatibility, but it would make us GAE
>>>>>>> compatible (as far as I can tell, this is the only think in the way of
>>>>>>> being GAE compatible).
>>>>>>>
>>>>>>> It is interesting that the ServletRequest API does not use the
>>>>>>> InetAddress class either, but also just uses Strings.
>>>>>>>
>>>>>>> Thoughts?
>>>>>>>
>>>>>>> - Les
>>>>>>>
>>>>>>> On Mon, Nov 23, 2009 at 10:14 AM, Kalle Korhonen
>>>>>>> <[email protected]>  wrote:
>>>>>>>
>>>>>>>
>>>>>>>> Restricted - as in it's not in the the whitelist
>>>>>>>> (http://code.google.com/appengine/docs/java/jrewhitelist.html). There
>>>>>>>> are only certain classes that are allowed to be loaded. We could make
>>>>>>>> using InetAddress optional. But Google is known to make changes to the
>>>>>>>> whitelist, allowing more classes upon request.
>>>>>>>>
>>>>>>>> Kalle
>>>>>>>>
>>>>>>>>
>>>>>>>> On Mon, Nov 23, 2009 at 6:47 AM, Les Hazlewood<[email protected]>
>>>>>>>>  wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>> Restricted in what way?  Not instantiated?  Or the class is not even
>>>>>>>>> allowed to be loaded by the classloader?
>>>>>>>>>
>>>>>>>>> Do you have a stack trace you could send along?
>>>>>>>>>
>>>>>>>>> The InetAddress is not strictly mandatory - it is there as a security
>>>>>>>>> feature to retain the IP from where the end-user initiates their
>>>>>>>>> application interaction so it is available easily to the application,
>>>>>>>>> something valuable to many apps.  Something might be able to be done
>>>>>>>>> to make this work in App Engine easily, but I'm not sure what without
>>>>>>>>> knowledge of what is going wrong.  Please also feel free to share any
>>>>>>>>> links to resources about how to deal with these restrictions if you
>>>>>>>>> have them.
>>>>>>>>>
>>>>>>>>> Cheers,
>>>>>>>>>
>>>>>>>>> Les
>>>>>>>>>
>>>>>>>>> On Sun, Nov 22, 2009 at 11:44 PM, rwilson<[email protected]>
>>>>>>>>>  wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>> Has anyone successfully integrated shiro with google app engine.
>>>>>>>>>> java.net.InetAddress is a restricted class.
>>>>>>>>>> Any suggestions?
>>>>>>>>>>
>>>>>>>>>> Thanks,
>>>>>>>>>> Roy
>>>>>>>>>> --
>>>>>>>>>> View this message in context:
>>>>>>>>>> http://n2.nabble.com/google-app-engine-tp4049065p4049065.html
>>>>>>>>>> Sent from the Shiro User mailing list archive at Nabble.com.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>
>>>
>>
>

Reply via email to